欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对现有压缩域图像增强算法在提高图像对比度时,存在不能很好地增强图像细节及保持色彩信息的局限性,提出一种新的基于Retinex理论的DCT压缩域图像增强算法.该算法以Retinex理论为基础,将DCT系数分为入射分量（DC系数）和反射分量（AC系数）,通过对DC系数进行动态范围调整,对AC系数进行细节增强调整,并使用阈值方法抑制块状效应,由此实现对压缩域图像的增强.实验结果表明,与传统的Retinex增强算法及DCT压缩域增强算法相比,该算法具有更好的细节增强及色彩保持效果,且能较好地抑制块状效应.
英文摘要：
Existing image enhancement algorithms in the compressed domain can not preserve the details and color information when enhancing the contrast of images.They mostly treat the DCT coefficients uniformly using the same strategy,or can not inhibit artifacts if treating the coefficients differently.A new image enhancement algorithm in the DCT compressed domain based on Retinex theory is proposed.The algorithm divides DCT coefficients into the illumination component（DC coefficients） and the reflection component（AC coefficients） according to Retinex theory.The DC coefficients are adjusted to compress the image dynamic range by mapping the illumination component to an ideal range,using two simple but powerful functions.The AC coefficients which represent reflection component are adjusted to enhance the local details with a new definition of spectral content measure.The fact that perception of image detail information for peopleis based on the ratio of high and low frequency in human visual system leads to the new definition.The block artifacts will be suppressed using a block decomposing strategy with a threshold automatically according to the characteristics of the images to be enhanced.Compared with traditional Retinex and DCT enhancement algorithms,experiment results show the proposed algorithm＇s efficiency in details enhancement and color preserving,also with artifacts compressed.
同期刊论文项目
同项目期刊论文
期刊信息
位置：